Methacholine challenge testing is common for assessing the pharmacodynamic properties of anti-asthma drugs. In order to design studies and to interpret published studies, sample size calculations are essential. Unfortunately, wrong sample sizes were previously reported in the literature. We present correct sample sizes required for the comparison of two treatments based on methacholine challenge testing in a crossover study.Formulas for sample size calculations and the resulting number of subjects required for a specified power are presented for studies designed to show a difference as well as for equivalence and non-inferiority studies.A much larger sample size is required for methacholine challenge testing than previously reported.